I tried this myself in Word on OS X with the default settings and got an exact match with respect to spacing, line breaks and so on. The question now is who did it and why did they think they could get away with such an awful job!
From the evidence that contradicts this claim: “It’s possible that the original was generated on a typewriter with a proportional width Times New Roman font wheel or ball (available at the time), with the same common margin settings as Word uses by default.” – that’s complete rubbish, becase a font wheel or ball would not be able to do kerning (whereby some letters are printed closer to each other than others) – a feature built in to modern word processors.
Correct me if I’m wrong, but don’t military organisations mandate monospace fonts for anything other than material intended for public consumption?
If you redacted something in a document that used proportional fonts, an attacker could measure the space the blanked-out words covers, and infer which combinations of letters are possible. By running it through a dictionary, you could narrow down the possible words by quite a bit. Throw in a bit more context, and it is often possible to reconstruct the original. Use of monospace fonts is a security measure.
Google is our friend. With five minutes of searching, it becomes clear that IBM was producing typewriters with proportional fonts for over a decade prior to the supposed creation of the Bush documents. That doesn’t mean we’re not looking at forgeries… it just means that folks have to stop all of that knee-jerking and let actual experts in the field analyze things.
Roger: I don’t think it’s knee-jerking to find it incredibly suspicious that retyping the memo in Microsoft Word using the default settings for a new document produces an identical document. Even with an incredibly expensive proportional font typewriter (something that would never be used for a regular military memo) you won’t get kerning, the highly complex algorithm used by modern word processors which adjusts the distance between letters based on what those letters are. I’d be amazed if it turned out that those memos weren’t forgeries produced using Microsoft Word.
Even if you don’t read the whole thing, scroll down far enough to look at the graphic comparing fonts at different sizes.
The problem appears to have been, as with so many failings in analysis both liberal and conservative, pulling back to just the right point that enough details are obscured that you can use the evidence to support they hypothesis you really want it to “prove”.
I’ve read the Daily KOS thread, and I still don’t buy it. What are the chances of a memo written on a typewriter in 1972 turning out virtually /identical/ to one typed in to Word using the default size/margin settings in 2004? To believe that this is a coincidence, even based on the fact that Times New Roman is a well established typeface with a history dating back to the 30s, seems naive at best. Show me a confirmed document from 1972 that looks identical when typed in Word today using the default settings and I’ll be a lot more receptive. I don’t doubt for a second that Bush used special connections to avoid Vietnam but these particular documents couldn’t be more dubious if they tried.
But Simon the point – at least as I understood it – is that it isn’t identical when viewed actual size. It’s only when reduced that the over-riding similarities come to the forefront. I thought the graphic of the two very different typefaces overlaid made that pretty clear.
At this point I’m looking purely at the technical criticisms being made against the document and I don’t think they hold water. Leave the content aside for a second. Typographic design lays out a set of rules which were being emulated by both the typewriter manufacturers and the word processing software manufacturers. We shouldn’t be at all surprised by gross similarities. Paper sizes haven’t changed. Standard business practices for letter writing (e.g. proper margins) probably haven’t either.
Has anyone tried this with a known document of the same era? That’s a pretty simple test someone ought to do. How much alike does the a new Word version look to a document no one doubts is 30+ years old? I’m betting that with minimal effort it would be possible to make something at least as similar as the supposed proofs that this Bush document is fake.
This doesn’t mean the document isn’t fake; I’m sure that a typewriter of that era is available somewhere and a fake could have been made without the help of computer software. It’s the unscientific approach to the whole forensic problem which annoys me.
“The probability that any technology in existence in 1972 would be capable of producing a document that is nearly pixel-compatible with Microsoft’s Times New Roman font and the formatting of Microsoft Word, and that such technology was in casual use at the Texas Air National Guard, is so vanishingly small as to be indistinguishable from zero.”
Priceless.
I tried this myself in Word on OS X with the default settings and got an exact match with respect to spacing, line breaks and so on. The question now is who did it and why did they think they could get away with such an awful job!
Oops, hadn’t followed the link when I commented. My last comment makes more sense if you’ve read this: http://littlegreenfootballs.com/weblog/?entry=12526_Bush_Guard_Documents-_Forged
Hang on. There’s also evidence that contradicts this claim. Some crucial details don’t match Word and could in fact be from a 70s era typewriter.
From the evidence that contradicts this claim: “It’s possible that the original was generated on a typewriter with a proportional width Times New Roman font wheel or ball (available at the time), with the same common margin settings as Word uses by default.” – that’s complete rubbish, becase a font wheel or ball would not be able to do kerning (whereby some letters are printed closer to each other than others) – a feature built in to modern word processors.
Correct me if I’m wrong, but don’t military organisations mandate monospace fonts for anything other than material intended for public consumption?
If you redacted something in a document that used proportional fonts, an attacker could measure the space the blanked-out words covers, and infer which combinations of letters are possible. By running it through a dictionary, you could narrow down the possible words by quite a bit. Throw in a bit more context, and it is often possible to reconstruct the original. Use of monospace fonts is a security measure.
sloppy work.
Simon: “that’s complete rubbish…”
Google is our friend. With five minutes of searching, it becomes clear that IBM was producing typewriters with proportional fonts for over a decade prior to the supposed creation of the Bush documents. That doesn’t mean we’re not looking at forgeries… it just means that folks have to stop all of that knee-jerking and let actual experts in the field analyze things.
Roger: I don’t think it’s knee-jerking to find it incredibly suspicious that retyping the memo in Microsoft Word using the default settings for a new document produces an identical document. Even with an incredibly expensive proportional font typewriter (something that would never be used for a regular military memo) you won’t get kerning, the highly complex algorithm used by modern word processors which adjusts the distance between letters based on what those letters are. I’d be amazed if it turned out that those memos weren’t forgeries produced using Microsoft Word.
Let the experts decide
Simon, I think you need to be amazed now: http://www.dailykos.com/story/2004/9/10/213416/348
Even if you don’t read the whole thing, scroll down far enough to look at the graphic comparing fonts at different sizes.
The problem appears to have been, as with so many failings in analysis both liberal and conservative, pulling back to just the right point that enough details are obscured that you can use the evidence to support they hypothesis you really want it to “prove”.
I’ve read the Daily KOS thread, and I still don’t buy it. What are the chances of a memo written on a typewriter in 1972 turning out virtually /identical/ to one typed in to Word using the default size/margin settings in 2004? To believe that this is a coincidence, even based on the fact that Times New Roman is a well established typeface with a history dating back to the 30s, seems naive at best. Show me a confirmed document from 1972 that looks identical when typed in Word today using the default settings and I’ll be a lot more receptive. I don’t doubt for a second that Bush used special connections to avoid Vietnam but these particular documents couldn’t be more dubious if they tried.
But Simon the point – at least as I understood it – is that it isn’t identical when viewed actual size. It’s only when reduced that the over-riding similarities come to the forefront. I thought the graphic of the two very different typefaces overlaid made that pretty clear.
At this point I’m looking purely at the technical criticisms being made against the document and I don’t think they hold water. Leave the content aside for a second. Typographic design lays out a set of rules which were being emulated by both the typewriter manufacturers and the word processing software manufacturers. We shouldn’t be at all surprised by gross similarities. Paper sizes haven’t changed. Standard business practices for letter writing (e.g. proper margins) probably haven’t either.
Has anyone tried this with a known document of the same era? That’s a pretty simple test someone ought to do. How much alike does the a new Word version look to a document no one doubts is 30+ years old? I’m betting that with minimal effort it would be possible to make something at least as similar as the supposed proofs that this Bush document is fake.
This doesn’t mean the document isn’t fake; I’m sure that a typewriter of that era is available somewhere and a fake could have been made without the help of computer software. It’s the unscientific approach to the whole forensic problem which annoys me.
By the way, if all this has awoken an interest in typography in anyone, I recommend the book Stop Stealing Sheep & Find Out How Type Works (http://www.amazon.com/exec/obidos/ASIN/0201703394/).
http://homepage.mac.com/cfj/expert.htm
“The probability that any technology in existence in 1972 would be capable of producing a document that is nearly pixel-compatible with Microsoft’s Times New Roman font and the formatting of Microsoft Word, and that such technology was in casual use at the Texas Air National Guard, is so vanishingly small as to be indistinguishable from zero.”
And that’s from a Microsoft typography expert.